In our upcoming report on Big Social, the social part of Big Data, we describe some new practices emerging from working with social media data. One of these practices is relating unrelated things and finding actionable data in these new relations. An ambitious Facebook data-collection effort is turning up such unlikely patterns in members’ interests.
One of these unlikely patterns retrieved from the Facebook data is that cyclists are more likely to own tablet computers. These may seem like random associations. “But there are patterns in all that randomness,” Ian Lurie, founder and CEO of Internet marketing agency Portent writes. These patterns open up new doors for marketers to find their way into potential customers’ lives.
Lurie’s work shows how two seemingly unrelated interests can be connected by a common thread, a concept he calls random affinity. He writes the following in his blogpost on the concept:
Everyone appears random. But there are patterns in all that randomness.
And from that chaos was born my latest project: The Idea Graph. I’m starting with Facebook data, later expanding to other social networks. The Graph maps out unexpected relationships, not between people (the social graph), but between ideas. It’s a social graph for topics.
To grow it, though, I need your help. I need a huge database of ‘likes’, grouped by Facebook user, to make this work. I’m only using the data anonymously. And I’ll be releasing the first dataset at MozCon, for free.
The result will be a list of topics and their distance – the likelihood that someone who likes that topic will like another particular topic.
Luring is looking for 16,000 volunteers to help build a data set that will map patterns of interest among Facebook members. You can participatie by adding a app to Facebook that anonymously tracks likes.
Random affinities could be another way to attract and keep your long tail audience. For me it’s just a good example of some new marketing practices coming from using with (big) social data. More info on the use and value of this idea graph can be found here.